U.S. patent application number 12/373402 was filed with the patent office on 2009-12-10 for photoelectric conversion device and imaging device.
Invention is credited to Kohji Shinomiya.
Application Number | 20090302360 12/373402 |
Document ID | / |
Family ID | 38956628 |
Filed Date | 2009-12-10 |
United States Patent
Application |
20090302360 |
Kind Code |
A1 |
Shinomiya; Kohji |
December 10, 2009 |
PHOTOELECTRIC CONVERSION DEVICE AND IMAGING DEVICE
Abstract
A photoelectric conversion device adopts the structure
reflecting the finding that color separation by the photoelectric
conversion, which utilizes the difference of the PN junction depth
of a semiconductor region, has the strong tendency that separation
of a B signal is easy but separation of a G signal and an R signal
becomes imperfect. That is, to cope with the tendency of the
imperfect color separation of a G signal and an R signal, PN
junction surfaces (JNC_B, JNC_R) of two photodiodes (PDs) for R
light and B light are superimposed in the depth direction, and PD
to G light is arranged independently. Accordingly, the color
separation property of each RGB light wavelength band can be
improved, the occupying area can be reduced compared with the case
where each PD of RGB light is dispersed in the plane direction, and
simplification of the semiconductor layer structure can be
realized.
Inventors: |
Shinomiya; Kohji; (Tokyo,
JP) |
Correspondence
Address: |
MATTINGLY & MALUR, P.C.
1800 DIAGONAL ROAD, SUITE 370
ALEXANDRIA
VA
22314
US
|
Family ID: |
38956628 |
Appl. No.: |
12/373402 |
Filed: |
July 21, 2006 |
PCT Filed: |
July 21, 2006 |
PCT NO: |
PCT/JP2006/314472 |
371 Date: |
February 13, 2009 |
Current U.S.
Class: |
257/292 ;
257/E27.135 |
Current CPC
Class: |
H01L 27/14647 20130101;
H01L 27/14627 20130101; H01L 27/14623 20130101; H01L 31/035236
20130101 |
Class at
Publication: |
257/292 ;
257/E27.135 |
International
Class: |
H01L 27/146 20060101
H01L027/146 |
Claims
1. A Photoelectric conversion device comprising: a first
semiconductor region having a first conductive type; a second
semiconductor region and a third semiconductor region, both having
a second conductive type and arranged in the first semiconductor
region; a fourth semiconductor region having the first conductive
type and arranged in the third semiconductor region; and a fifth
semiconductor region having the second conductive type and arranged
in the fourth semiconductor region, wherein the first semiconductor
region and the second semiconductor region constitute a first
photodiode, and a junction surface between the first semiconductor
region constituting an anode of the first photodiode and the second
semiconductor region constituting a cathode of the first photodiode
has a first depth for photoelectric conversion to light existing in
a medium wavelength band and entering from a surface of the first
semiconductor region, wherein the fourth semiconductor region and
the third semiconductor region constitute a second photodiode, and
a junction surface between the fourth semiconductor region
constituting an anode of the second photodiode and the third
semiconductor region constituting a cathode of the second
photodiode has a second depth for photoelectric conversion to light
existing in a long wavelength band and entering from a surface of
the first semiconductor region, and wherein the fourth
semiconductor region and the fifth semiconductor region constitute
a third photodiode, and a junction surface between the fourth
semiconductor region constituting an anode of the third photodiode
and the fifth semiconductor region constituting a cathode of the
third photodiode has a third depth for photoelectric conversion to
light existing in a short wavelength band and entering from a
surface of the first semiconductor region.
2. The photoelectric conversion device according to claim 1,
wherein the superimposed second photodiode and third photodiode and
the first photodiode are arranged in a matrix in the shape of a
checkered pattern.
3. The photoelectric conversion device according to claim 1,
wherein the light in the long wavelength band is red light, the
light in the medium wavelength band is green light, and the light
in the short wavelength band is blue light.
4. The photoelectric conversion device according to claim 1,
wherein the first to the fifth semiconductor region have, over each
surface, a high-concentration impurity layer of the first
conductive type, and wherein the high-concentration impurity layer
couples electrically the first semiconductor region and the fourth
semiconductor region.
5. The photoelectric conversion device according to claim 1,
further comprising: a first transfer MOS transistor of which one of
a source and a drain is served by the second semiconductor region
and of which the other one of the source and the drain is formed by
a semiconductor region of the second conductive type provided in
the first semiconductor region; a second transfer MOS transistor of
which one of a source and a drain is served by the third
semiconductor region and of which the other one of the source and
the drain is formed by a semiconductor region of the second
conductive type provided in the first semiconductor region; and a
third transfer MOS transistor of which one of a source and a drain
is served by the fifth semiconductor region and of which the other
one of the source and the drain is formed by a semiconductor region
of the second conductive type provided in the first semiconductor
region, wherein a charge accumulation-output unit is provided at
each of the first to the third photodiode, the charge
accumulation-output unit being operable to accumulate, via each
transfer MOS transistor, charge information produced by a current
which is induced by photoelectric conversion and flows in each
junction surface, and operable to output the accumulated charge
information.
6. The photoelectric conversion device according to claim 1,
further comprising: a first transfer MOS transistor of which one of
a source and a drain is served by the second semiconductor region
and of which the other one of the source and the drain is formed by
a semiconductor region of the second conductive type provided in
the first semiconductor region; a second transfer MOS transistor of
which one of a source and a drain is served by the third
semiconductor region and of which the other one of the source and
the drain is formed by a semiconductor region of the second
conductive type provided in the first semiconductor region; and a
third transfer MOS transistor of which one of a source and a drain
is served by the fifth semiconductor region and of which the other
one of the source and the drain is formed by a semiconductor region
of the second conductive type provided in the first semiconductor
region, wherein a charge accumulation-output unit is provided in
common at the first photodiode and the second photodiode and
provided exclusively at the third photodiode, the charge
accumulation-output unit being operable to accumulate, via each
transfer MOS transistor, charge information produced by a current
which is induced by photoelectric conversion and flows in each
junction surface, and operable to output the accumulated charge
information.
7. The photoelectric conversion device according to claim 1,
further comprising: a first transfer MOS transistor of which one of
a source and a drain is served by the second semiconductor region
and of which the other one of the source and the drain is formed by
a semiconductor region of the second conductive type provided in
the first semiconductor region; a second transfer MOS transistor of
which one of a source and a drain is served by the third
semiconductor region and of which the other one of the source and
the drain is formed by a semiconductor region of the second
conductive type provided in the first semiconductor region; and a
third transfer MOS transistor of which one of a source and a drain
is served by the fifth semiconductor region and of which the other
one of the source and the drain is formed by a semiconductor region
of the second conductive type provided in the first semiconductor
region, wherein a charge accumulation-output unit is provided in
common at the first to the third photodiode, the charge
accumulation-output unit being operable to accumulate, via each
transfer MOS transistor, charge information produced by a current
which is induced by photoelectric conversion and flows in each
junction surface, and operable to output the accumulated charge
information.
8. The photoelectric conversion device according to claim 5,
wherein all or a part of the first to the third transfer MOS
transistor are bulk MOS transistors, and wherein each of the bulk
MOS transistors has, at a boundary surface under a gate, an
impurity region of higher impurity concentration than a channel
forming layer.
9. The photoelectric conversion device according to claim 5,
wherein the charge accumulation-output unit includes: a source
follower output transistor of which a gate is coupled to the other
one of the source and the drain of each transfer MOS transistor;
and a reset MOS transistor operable to charge selectively a path
from the gate of the source follower output transistor to a cathode
of the corresponding photodiode.
10. The photoelectric conversion device according to claim 9,
wherein all or a part of the first to the third transfer MOS
transistor, the source follower output transistor, and the reset
MOS transistor are bulk MOS transistors, and wherein each of the
bulk MOS transistors has, at a boundary surface under a gate, an
impurity region of higher impurity concentration than a channel
forming layer.
11. The photoelectric conversion device according to claim 1,
further comprising: a light shielding film with openings provided
in the respective upper parts of the second semiconductor region
and the fifth semiconductor region; a concave lens formed by
translucent material and arranged at each of the openings; and a
convex lens formed by translucent material and arranged over the
concave lens.
12. The photoelectric conversion device according to claim 1,
further comprising: a light shielding film with openings provided
in the respective upper parts of the second semiconductor region
and the fifth semiconductor region, wherein the second to the fifth
semiconductor regions are formed so as to have the first to the
third depth in the direction of a light diffracted outwards at a
periphery portion of the opening, similarly as in the direction of
a light entering substantially perpendicular to the opening.
13. A solid-state imaging device comprising: a first photodiode; a
second photodiode; and a third photodiode, wherein the first
photodiode, the second photodiode, and the third photodiode
constitute the photoelectric conversion device according to claim
1, and wherein the first photodiode and the superimposed second
photodiode and third photodiode are arranged in a matrix on a
semiconductor substrate.
14. An imaging device comprising: a solid-state imaging device
according to claim 13; an analog front-end unit operable to
digitize an image signal produced by the solid-state imaging
device; and a digital signal processing processor operable to
generate image data by performing digital signal processing to
output data of the analog front-end unit.
15. A photoelectric conversion device comprising: a first
semiconductor region having a first conductive type; a second
semiconductor region and a third semiconductor region, both having
a second conductive type and arranged in the first semiconductor
region; and a fourth semiconductor region having the second
conductive type, arranged in the first semiconductor region, and
partly overlapped with the third semiconductor region in the depth
direction, wherein the first semiconductor region and the second
semiconductor region constitute a first photodiode, and a junction
surface between the first semiconductor region constituting an
anode of the first photodiode and the second semiconductor region
constituting a cathode of the first photodiode has a first depth
for photoelectric conversion to light existing in a medium
wavelength band and entering from a surface of the first
semiconductor region, wherein the first semiconductor region and
the third semiconductor region constitute a second photodiode, and
a junction surface between the first semiconductor region
constituting an anode of the second photodiode and the third
semiconductor region constituting a cathode of the second
photodiode has a second depth for photoelectric conversion to light
existing in a long wavelength band and entering from a surface of
the first semiconductor region, and wherein the first semiconductor
region and the fourth semiconductor region constitute a third
photodiode, and a junction surface between the first semiconductor
region constituting an anode of the third photodiode and the fourth
semiconductor region constituting a cathode of the third photodiode
has a third depth for photoelectric conversion to light existing in
a short wavelength band and entering from a surface of the first
semiconductor region.
Description
TECHNICAL FIELD
[0001] The present invention relates to a photoelectric conversion
device used for detection of a color image, and to technology which
is effective when applied to an image input device, such as a
camera or a solid-state imaging device, for example.
BACKGROUND ART
[0002] In the past, a color filter (written as CF in the following)
to pass the light of red, green, and blue (written as RGB in the
following) is inserted in a light path to a pixel, and
photoelectric conversion is performed with a photodiode (written as
PD in the following) formed in the pixel. Consequently, a color
image is produced with the constitution in which a red, green, and
blue (RGB) electrical signal are independently obtained for every
pixel respectively.
[0003] Since the system which performs a color separation by the
RGB-CF requires a pixel dedicated for every color, three RGB pixels
are necessary in order to obtain a color electrical signal in a set
of at least three pixels, such as RGB.
[0004] Since it is necessary to make light transmit an RGB-CF,
light energy is converted into heat by the RGB-CF and decreases,
and the transmitted light energy which has decreased is irradiated
to PD. Therefore, there arises a problem that the photoelectron
after photoelectric conversion decreases and an output chrominance
signal is reduced, as well as a problem that the heat generation by
the RGB-CF varies the spectroscopic property of the RGB-CF. Another
problem is decrease of light use efficiency, due to the thickness
of the RGB-CF layer which narrows the angle of available incident
light. Since a pigment is used for the material of an RGB-CF, in
order to protect a wafer manufacturing process line from
contamination by the heavy metal included in the pigment, an
exclusive RGB-CF manufacturing line, separated from the wafer
manufacturing process line, needs to be provided.
[0005] As a means to solve these problems, Patent Document 1 and
Patent Document 2 propose a system in which a color electrical
signal, such as RGB, is obtained without using an optical color
filter, such as an RGB-CF. Patent Document 1 proposes a color image
sensing device in which PN junctions of plural numbers of depth are
constituted, by laminating several layers of semiconductor region
with alternately different conductive types; accordingly,
photoelectric conversion is performed for light in different plural
wavelength band regions by the PN junctions of plural numbers of
depth, respectively.
[0006] Patent Document 2 proposes a color image sensing device in
which, in order to obtain RGB chrominance signals by a single pixel
without using an RGB-CF, PN junctions of plural numbers of depth
are constituted, by laminating several layers of semiconductor
region with alternately different conductive types; accordingly
photoelectric conversion is performed for light in different plural
wavelength band regions by the PN junctions of plural numbers of
depth, respectively.
[0007] Patent Document 1 and Patent Document 2 both propose
elements in which PN junctions of plural numbers of depth are
constituted, by laminating several layers of semiconductor region
with alternately different conductive types, accordingly
photoelectric conversion is performed for light in different plural
wavelength band regions by the PN junctions of plural numbers of
depth, respectively. According to the elements, a semiconductor is
irradiated without using an RGB-CF and there is no attenuation of
the light in the RGB-CF when compared with an RGB-CF system,
therefore, as one of features, the effective utilization efficiency
of light is high. It is not necessary to constitute three pixels
independently in two dimensions for every RGB, and three
chrominance signals of RGB can be obtained independently from one
pixel. Therefore, it is possible to reduce the PD occupied area of
a pixel to one third, compared with an RGB-CF system, and it is
also possible to attain about three times better resolution than
the RGB-CF system, and about three times higher sensitivity than
the RGB-CF system.
[0008] [Patent Document 1] Japanese Unexamined patent Publication
No. Sho 61 (1986)-187282.
[0009] [Patent Document 2] Japanese Unexamined patent Publication
No. 2003-298038.
DISCLOSURE OF THE INVENTION
[0010] Problems that the Invention is to Solve
[0011] In Patent Document 1, the PN junctions of plural numbers of
depth are constituted by laminating plural layers of semiconductor
region with alternately different conductive types in the depth
direction of the semiconductor, and each semiconductor layer of
each PD is shared alternately in the depth direction. Therefore,
there are shortcomings that it is difficult to get signal electrons
of each PD independently, because the signal electrons which are
photoelectrically converted in each PD may be influenced mutually.
Furthermore, there is a problem that a leakage current generated in
each PD flows into the other PDs, inducing an error in chrominance
signals, such as RGB.
[0012] Patent Document 2 makes improvements by providing a
semiconductor region which aims at separation between each PD so
that each PD may be able to operate independently, and the
structure of eight semiconductor regions is formed by laminating
seven junctions of PNPNPNPN in the depth direction of the
semiconductor. Therefore, shortcomings are that the structure is so
complicated that actually manufacturing as a product is extremely
difficult.
[0013] Examination by the present inventors has clarified that when
an incident light to PD arranged directly under an opening of a
light shielding film is diffracted by the opening periphery, the
direction of the light becomes slanting and the light path length
becomes longer, compared with a perpendicular light. This fact
means that a light of different wavelength may be detected by the
same PD, leading to decrease of a wavelength separation accuracy or
a color separation accuracy. It has been also clarified that an
undesirable dark current flows in the surface portion of a
semiconductor region, on the ground of the contamination on the
surface of the semiconductor region due to semiconductor process,
leading to decrease of photoelectric conversion accuracy.
[0014] The present invention has been made in view of the above
circumstances and provides a photoelectric conversion device which
adopts the structure of PN junctions of plural numbers of depth,
constituted by laminating alternately semiconductor regions with
different conductive types in the depth direction of the
semiconductor, without using an optical color filter, such as an
RGB-CF. Accordingly, the photoelectric conversion device can get a
less-noisy electrical signal by improving the color separation
property of each light wavelength band and can be actually
manufactured as a product by simplifying the structure as much as
possible. The present invention also provides an imaging device
employing such a photoelectric conversion device.
[0015] The other purposes and the new feature of the present
invention will become clear from the description of the present
specification and the accompanying drawings.
Means for Solving the Problems
[0016] The following simply explains an outline of typical one of
the inventions disclosed by the present application.
[0017] A photoelectric conversion device related to the present
invention adopts the structure reflecting the finding that color
separation by the photoelectric conversion, which utilizes the
difference of the PN junction depth of a semiconductor region, has
the strong tendency that separation of a B signal is easy but
separation of a G signal and an R signal becomes imperfect. That
is, to cope with the tendency of the imperfect color separation of
a G signal and an R signal, PN junction surfaces of two PDs for R
light and B light are superimposed in the depth direction, and PD
for G light is arranged independently. The signal obtained from the
shallowest PD for B light provides a converted signal of B light
with high accuracy. The signal obtained from the next deeper PD for
G light includes the converted signal of B light also. Therefore,
the converted signal of G light can be obtained with high accuracy
by subtracting the converted signal obtained by PD of B light from
the converted signal obtained by PD of G light, in a latter stage
circuit. Similarly, the signal obtained by the deepest PD for R
light includes the converted signal of B light and the converted
signal of G light. Therefore, the converted signal of R light can
be obtained with high accuracy by subtracting, from the converted
signal obtained by PD of R light, the converted signal obtained by
PD of B light and further subtracting the signal of G light
produced by the subtraction in the latter stage circuit.
Accordingly, the color separation property of each RGB light
wavelength band can be improved, the occupying area can be reduced,
compared with the case where each PD of RGB light is dispersed in
the plane direction, and simplification of the semiconductor layer
structure can be realized compared with the case where each PD of
RGB light is arranged in the depth direction.
[0018] The structure is explained in full detail. The photoelectric
conversion device according to the present invention includes a
first semiconductor region (1) of a first conductive type (for
example, P type), a second semiconductor region (2) and a third
semiconductor region (3) of a second conductive type (for example,
N type) which are formed in the first semiconductor region, a
fourth semiconductor region (4) of the first conductive type formed
in the third semiconductor region, and a fifth semiconductor region
(5) of the second conductive type formed in the fourth
semiconductor region. The first semiconductor region and the second
semiconductor region constitute a first photodiode (PDG). The
junction surface between the first semiconductor region,
constituting an anode of the first photodiode, and the second
semiconductor region, constituting a cathode of the first
photodiode, has a first depth (DP_G) for photoelectric conversion
to light existing in a medium wavelength band and entering from a
surface of the first semiconductor region. The fourth semiconductor
region and the third semiconductor region constitute a second
photodiode (PDR). The junction surface between the fourth
semiconductor region, constituting an anode of the second
photodiode, and the third semiconductor region, constituting a
cathode of the second photodiode, has a second depth (DP_R) for
photoelectric conversion to light existing in a long wavelength
band and entering from a surface of the first semiconductor region.
The fourth semiconductor region and the fifth semiconductor region
constitute a third photodiode (PDB). The junction surface between
the fourth semiconductor region, constituting an anode of the third
photodiode, and the fifth semiconductor region, constituting a
cathode of the third photodiode, has a third depth (DP_B) for
photoelectric conversion to light existing in a short wavelength
band and entering from a surface of the first semiconductor
region.
[0019] The light in the long wavelength band is red light, the
light in the medium wavelength band is green light, and the light
in the short wavelength band is blue light.
[0020] To cope with the color separation imperfection of a G signal
and an R signal when superimposing the PN junction surfaces of
three sorts of PDs corresponding to the light wavelengths of RGB in
the depth direction, a PD to G light which has a poorest color
separation performance is arranged independently, and the PN
junction surfaces to two sorts of PDs for R light and B light are
superimposed in the depth direction. Accordingly, reduction of the
occupying area is attained. It is effective to arrange the
superimposed second photodiode and third photodiode, and the first
photodiode in a matrix in the shape of a checkered pattern.
According to this arrangement, by assigning a pixel for every
photodiode in a planar arrangement of the photodiodes arranged in a
matrix, and by performing an interpolation color operation using
pixels which adjoin mutually in a vertical and a horizontal
direction, it is possible to detect an image with the same pixel
number as the element number of the array of the photodiodes
arranged in a matrix.
[0021] As one specific mode of the present invention, the first to
the fifth semiconductor region have, over each surface, a
high-concentration impurity layer of the first conductive type, and
the high-concentration impurity layer couples electrically the
first semiconductor region and the fourth semiconductor region. The
high-concentration impurity layer acts so as to pull in an
undesirable dark current, which flows in the surface portion of a
semiconductor region, on the ground of the contamination on the
surface of the semiconductor region due to semiconductor process,
to a common potential or a ground potential to which the anodes
(A_C) of the first to the third photodiode are coupled. Therefore,
the situation where such a dark current flows into the cathode and
reduces the photoelectric conversion accuracy can be
suppressed.
[0022] As another specific mode of the present invention, the
photoelectric conversion device includes: a first transfer MOS
transistor (M1) of which one of a source and a drain is served by
the second semiconductor region and of which the other one of the
source and the drain is formed by a semiconductor region of the
second conductive type provided in the first semiconductor region;
a second transfer MOS transistor (M11) of which one of a source and
a drain is served by the third semiconductor region and of which
the other one of the source and the drain is formed by a
semiconductor region of the second conductive type provided in the
first semiconductor region; a third transfer MOS transistor (M21)
of which one of a source and a drain is served by the fifth
semiconductor region and of which the other one of the source and
the drain is formed by a semiconductor region of the second
conductive type provided in the first semiconductor region; and a
charge accumulation-output unit which is provided at each of the
first to the third photodiode (ACCG, ACCR, ACCB) and accumulates,
via each transfer MOS transistor, charge information produced by a
current which is induced by photoelectric conversion and flows in
each junction surface, and outputs the accumulated charge
information. The imaging cycle using a photoelectric conversion
element is classified roughly into a reset cycle, an exposure
cycle, and a transfer cycle. In the reset cycle, the transfer MOS
transistor is turned on, and an initial charge is stored in the
charge accumulation-output unit and the cathode of the photodiode.
In the exposure cycle, the transfer MOS transistor is turned off,
and the photodiode is rendered to perform photoelectric conversion.
In the next transfer cycle, the transfer MOS transistor is turned
on, the photoelectron (electron produced by the photoelectric
conversion) accumulated in the cathode of the photodiode is
transferred to the charge accumulation-output unit, and after
turning the transfer MOS transistor off subsequently, a converted
signal is taken out from the charge accumulation-output unit. Since
the transfer MOS transistor is provided between the charge
accumulation-output unit and the cathode of the photodiode, it is
possible to suppress that the converted signal becomes unstable
under the influence of a noise, when taking out the converted
signal from the charge accumulation-output unit. In the
constitution in which the charge accumulation-output unit is
provided for each of the first to third photodiode, the detection
signals to each wavelength of R, G, and B can be outputted in
parallel, by parallel operation of the charge accumulation-output
unit.
[0023] The charge accumulation-output unit may be provided in
common at the first photodiode and the second photodiode (ACCRB),
and provided exclusively at the third photodiode (ACCG). The
occupied area for the case can be made smaller than the above
constitution.
[0024] The charge accumulation-output unit may be provided in
common at the first to the third photodiode (ACCRBG). The occupied
area for the case can be made still smaller.
[0025] As a yet specific mode, the charge accumulation-output unit
includes a source follower output transistor (M2, M12, M22, M32,
M42) of which a gate is coupled to the other one of the source and
the drain of the transfer MOS transistor, and a reset MOS
transistor (M4, M14, M24, M34, M44) which charges selectively a
path from the gate of the source follower output transistor to a
cathode of the corresponding PD. The reset MOS transistor is turned
on in the reset cycle, and turned off in the other cycles.
[0026] As a yet specific mode, all or a part of the first to the
third transfer MOS transistor, the source follower output
transistor, and the reset MOS transistor may employ a bulk MOS
transistor. The bulk MOS transistor has, at a boundary surface
under a gate, an impurity region of higher impurity concentration
than a channel forming layer. In the bulk MOS transistor, a channel
is not formed in the surface directly under the gate, and even if
undesirable contamination exists in the surface concerned, a
channel current cannot be easily influenced by the noise current
due to the contamination. Accordingly, a noise reduction effect is
expected.
[0027] As another specific mode of the present invention, the
photoelectric conversion device includes: a light shielding film
with openings provided in the respective upper parts of the second
semiconductor region and the fifth semiconductor region; a concave
lens (23) formed by translucent material and arranged at each of
the openings; and a convex lens (24) formed by the translucent
material and arranged over the concave lens. The convex lens
converges the incident light to a photodiode, and improves the
condensing of light. The concave lens can change the condensed
light to a parallel light, and is able to render the light enter
substantially perpendicular to the photodiode. Accordingly,
excellent color separation performance can be obtained as well as
the condensing of light.
[0028] As another specific mode of the present invention, the
photoelectric conversion device includes a light shielding film
(22) with openings (21) provided in the respective upper parts of
the second semiconductor region and the fifth semiconductor region.
The second to the fifth semiconductor regions are formed so as to
have the first to the third depth in the direction of a light
diffracted outwards at a periphery portion of the opening,
similarly as in the direction of a light entering substantially
perpendicularly to the opening. Even when the incident light of a
photodiode is diffracted at the opening periphery and the direction
of the light becomes slanting, the light path length cannot change
greatly compared with the perpendicular light; therefore, the
accuracy deterioration of the wavelength separation or the color
separation in one photodiode can be suppressed.
[0029] A solid-state imaging device includes the first photodiode,
the superimposed second photodiode and third photodiode, which
constitute a photoelectric conversion device and which are arranged
in the shape of an array on one semiconductor substrate.
[0030] The imaging device includes: a solid-state imaging device;
an analog front-end unit which digitizes an image signal produced
by the solid-state imaging device; and a digital signal processing
processor which generates image data by performing digital signal
processing to output data of the analog front-end unit.
[0031] A photoelectric conversion device related to another
structure of the present invention includes: a first semiconductor
region (1) having a first conductive type; a second semiconductor
region (2) and a third semiconductor region (3A), both having a
second conductive type and arranged in the first semiconductor
region; and a fourth semiconductor region (5A) having the second
conductive type, arranged in the first semiconductor region, and
partly overlapped with the third semiconductor region in the depth
direction. The first semiconductor region and the second
semiconductor region constitute a first photodiode, and a junction
surface between the first semiconductor region, constituting an
anode of the first photodiode, and the second semiconductor region,
constituting a cathode of the first photodiode, has a first depth
for photoelectric conversion to light existing in a medium
wavelength band and entering from a surface of the first
semiconductor region. The first semiconductor region and the third
semiconductor region constitute a second photodiode, and a junction
surface between the first semiconductor region, constituting an
anode of the second photodiode, and the third semiconductor region,
constituting a cathode of the second photodiode, has a second depth
for photoelectric conversion to light existing in a long wavelength
band and entering from a surface of the first semiconductor region.
The first semiconductor region and the fourth semiconductor region
constitute a third photodiode, and a junction surface between the
first semiconductor region, constituting an anode of the third
photodiode, and the fourth semiconductor region, constituting a
cathode of the third photodiode, has a third depth for
photoelectric conversion to light existing in a short wavelength
band and entering from a surface of the first semiconductor region.
The present structure is different from the photoelectric
conversion device described above in the point that the third
semiconductor region has a part embedded in the first semiconductor
region in a manner that both surfaces of the part touch with the
first semiconductor region. The present structure has the similar
function and effect with the photoelectric conversion device
described above.
EFFECTS OF THE INVENTION
[0032] The following explains briefly the effect acquired by the
typical one of the inventions disclosed by the present
application.
[0033] Namely, it is possible to simplify the structure of PN
junctions of plural numbers of depth constituted by laminating
alternately semiconductor regions with different conductive types
in the depth direction of the semiconductor, without using optical
color filters, such as an RGB-CF, and it is also possible to get a
less-noisy electrical signal by improving the color separation
property of each light wavelength band.
BRIEF DESCRIPTION OF THE DRAWINGS
[0034] FIG. 1 is an outline sectional view illustrating a
photoelectric conversion element which constitutes a solid-state
imaging device;
[0035] FIG. 2 is an explanatory view schematically illustrating the
distance of absorption of light which enters substantially
perpendicularly to a Si semiconductor region surface and is
absorbed in Si, in contrast with the device cross-section structure
of FIG. 1;
[0036] FIG. 3 is an explanatory view illustrating a comparative
example in which each PD of RGB light is dispersively arranged in
the plane direction;
[0037] FIG. 4 is an explanatory view illustrating a comparative
example in which all of PDs of RGB light are arranged in the depth
direction;
[0038] FIG. 5 is a circuit diagram illustrating a green photodiode
and a detector circuit of a photoelectric conversion signal by the
green photodiode;
[0039] FIG. 6 is an outline view of plane layout illustrating
planar structure of the circuit of FIG. 5;
[0040] FIG. 7 is an outline view of plane layout illustrating
planar structure of a blue photodiode and a red photodiode and a
detector circuit for photoelectric conversion signals by the blue
photodiode and the red photodiode;
[0041] FIG. 8 is an outline view of plane layout pattern
illustrating the entire plane constitution of one photoelectric
conversion element constituted by the green photodiode of FIG. 6
and the blue photodiode and the red photodiode of FIG. 7;
[0042] FIG. 9 is an entire layout pattern illustrating one
photoelectric conversion element related to the comparative
example, corresponding to FIG. 3, in which a green photodiode, a
blue photodiode, and a red photodiode are dispersively arranged in
plane;
[0043] FIG. 10 illustrates a circuit diagram when employing a
charge accumulation-output unit ACCRB provided in common to a blue
photodiode and a red photodiode;
[0044] FIG. 11 is a layout pattern illustrating the planar
structure of the circuit of FIG. 10;
[0045] FIG. 12 is a circuit diagram when employing a charge
accumulation-output unit ACCRBG provided in common to a green
photodiode, a blue photodiode, and a red photodiode;
[0046] FIG. 13 is a layout pattern illustrating the planar
structure of the circuit of FIG. 12;
[0047] FIG. 14 is a vertical cross-sectional view illustrating a
still-more-detailed vertical section structure of a photoelectric
conversion device to which a transfer MOS transistor is added;
[0048] FIG. 15 is a vertical cross-sectional view illustrating
another outline cross-section structure of a photoelectric
conversion device which constitutes a solid-state imaging
device;
[0049] FIG. 16 is a vertical cross-sectional view illustrating yet
another outline cross-section structure of a photoelectric
conversion device which constitutes a solid-state imaging
device;
[0050] FIG. 17 is a vertical cross-sectional view illustrating
circular vertical section structure in both ends of a third to a
fifth semiconductor region, as another example of the shape of the
laminated semiconductor regions;
[0051] FIG. 18 is a vertical cross-sectional view illustrating
vertical section structure which combines a lens with a red light
photodiode and a blue light photodiode superimposed in the vertical
direction, as another example in which the lens is arranged in
front of a photoelectric conversion element;
[0052] FIG. 19 is a system configuration diagram illustrating an
imaging device using a photoelectric conversion device according to
the present invention;
[0053] FIG. 20 is a vertical cross-sectional view of a bulk MOS
transistor; and
[0054] FIG. 21 is an explanatory view illustrating photodiodes
arranged in array in the shape of a checkered pattern, and the mode
of an interpolation color operation.
EXPLANATION OF REFERENCE NUMERALS
[0055] 1: First semiconductor region of first conductive type (for
example, N type)
[0056] 2: Second semiconductor region of second conductive type
(for example, P type)
[0057] 3, 3A: Third semiconductor region of second conductive type
(for example, P type)
[0058] 4: Fourth semiconductor region of first conductive type
[0059] 5: Fifth semiconductor region of second conductive type
[0060] 5A: Fourth semiconductor region of second conductive
type
[0061] 6: High-concentration impurity layer
[0062] JNC_R: PN junction surface of red photodiode
[0063] JNC_G: PN junction surface of green photodiode
[0064] JNC_B: PN junction surface of blue photodiode
[0065] K_R: Cathode terminal of red photodiode
[0066] K_G: Cathode terminal of green photodiode
[0067] K_B: Cathode terminal of blue photodiode
[0068] A_C: Common anode terminal
[0069] M1, M11, M21: Transfer MOS transistor
[0070] M2, M12, M22, M32, M42: Source follower output MOS
transistor
[0071] M3, M13, M23, M33, M43: Selection MOS transistor
[0072] M4, M14, M24, M34, M44: Reset MOS transistor
[0073] ACCR, ACCG, ACCB: Charge accumulation-output unit
[0074] 10: Gate oxide layer
[0075] 12: Light shielding film
[0076] 20: Antireflection film
[0077] 21: Opening
[0078] 22: Light shielding film
[0079] 23: Concave lens
[0080] 24: Convex lens
[0081] 30: Solid-state imaging device
[0082] 34: CDS (Correlated double sampling circuit)
[0083] 35: GCA (Gain control amplifier)
[0084] 36: ADC (Analog-digital converter)
[0085] 38: DSP (Digital signal processing processor)
BEST MODE FOR CARRYING OUT THE INVENTION
[0086] FIG. 1 is a basic outline sectional view illustrating a
photoelectric conversion element which constitutes a solid-state
imaging device as a photoelectric conversion device. The imaging
device arranges many photoelectric conversion elements in the shape
of an array on one semiconductor silicone substrate.
[0087] It is assumed that the semiconductor silicone substrate is
of N.sup.+-type, for example, and an N.sup.--type area is
epitaxially grown on the semiconductor silicone substrate. In the
N.sup.--type area, many photoelectric conversion elements are
formed at a predetermined pitch in a matrix.
[0088] The photoelectric conversion device includes: a first
semiconductor region 1, having a first conductive type, for
example, a P type, and arranged in the N.sup.--type area; a second
semiconductor region 2 and a third semiconductor region 3, both
having a second conductive type, for example, an N type, and
arranged in the first semiconductor region; a fourth semiconductor
region 4, having the first conductive type and arranged in the
third semiconductor region; and a fifth semiconductor region 5,
having the second conductive type and arranged in the fourth
semiconductor region. The first semiconductor region 1 and the
second semiconductor region 2 constitute a green photodiode (a
first photodiode). The fourth semiconductor region 4 and the third
semiconductor region 3 constitute a red photodiode (a second
photodiode). The fourth semiconductor region 4 and the fifth
semiconductor region 5 constitute a blue photodiode (a third
photodiode). The symbol K_G denotes a cathode terminal of the green
photodiode. The symbol K_R denotes a cathode terminal of the red
photodiode. The symbol K_B denotes a cathode terminal of the blue
photodiode. The first semiconductor region 1 through the fifth
semiconductor region 5 have a high-concentration impurity layer (a
cap layer) 6 of a P type over each surface. The high-concentration
impurity layer 6 conducts electrically the first semiconductor
region 1 and the fourth semiconductor region 4, and is coupled to a
common anode terminal A_C of the green photodiode (PDG), the red
photodiode (PDR), and the blue photodiode (PDB).
[0089] A junction surface JNC_G between the first semiconductor
region 1, constituting the anode of the green photodiode, and the
second semiconductor region 2, constituting the cathode of the
green photodiode, has a first depth (DP_G) for photoelectric
conversion to light existing in a medium wavelength band (for
example, green light represented by the wavelength of 520 nm (G
light)) and entering from the surface of the first semiconductor
region 1. A junction surface JNC_R between the fourth semiconductor
region 4, constituting the anode of the red photodiode, and the
third semiconductor region 3, constituting cathode of the red
photodiode, has a second depth (DP_R) for photoelectric conversion
to light existing in a long wavelength band (for example, red light
represented by the wavelength of 660 nm (R light)) and entering
from the surface of the first semiconductor region 1. A junction
surface JNC_B between the fourth semiconductor region 4,
constituting the anode of the blue photodiode, and the fifth
semiconductor region 5, constituting the cathode of the blue
photodiode, has a third depth (DP_B) for photoelectric conversion
to light existing in a short wavelength band (for example, blue
light represented by the wavelength of 450 nm (B light)) and
entering from the surface of the first semiconductor region 1. For
example, when the depth of the first semiconductor region 1 is set
to 8.0 micrometers (.mu.m), the second depth (DP_R) is set to 4.0
.mu.m, the first depth (DP_G) is set to 2.0 .mu.m, and the third
depth (DP_B) is set to 0.5 .mu.m.
[0090] The green photodiode, the red photodiode, and the blue
photodiode constitute a photoelectric conversion element,
respectively. The red photodiode and the blue photodiode are
superimposed in the depth direction.
[0091] FIG. 2 illustrates schematically the distance of absorption
of light which enters substantially perpendicularly to a Si
semiconductor region surface and is absorbed in Si, in contrast
with the device cross-section structure of FIG. 1. According to the
example of FIG. 2, B light is absorbed up to about 2.0 .mu.m from
the Si semiconductor region surface, G light is absorbed up to
about 3.5 .mu.m from the Si semiconductor region surface, and R
light is absorbed up to about 5.5 .mu.m from the Si semiconductor
region surface. Therefore, if the PN junction depths of a
semiconductor region are set differently, the color separation by
photoelectric conversion can be performed, even if the PN junction
surfaces of the respective RGB light are superimposed in the depth
direction. To describe in further detail, the signal obtained from
the shallowest PD for B light provides a converted signal of B
light with high accuracy. The signal obtained from the next deeper
PD for G light includes the converted signal of B light also.
Therefore, the converted signal of G light can be obtained with
high accuracy by subtracting the converted signal obtained by PD of
B light from the converted signal obtained by PD of G light, in a
latter stage circuit. Similarly, the signal obtained by the deepest
PD for R light includes the converted signal of B light and the
converted signal of G light. Therefore, the converted signal of R
light can be obtained with high accuracy by subtracting, from the
converted signal obtained by PD of R light, the converted signal
obtained by PD of B light and further subtracting the signal of G
light produced by the subtraction in the latter stage circuit. As
clearly seen from FIG. 2, color separation by the photoelectric
conversion, which utilizes the difference of the PN junction depth,
has the strong tendency that separation of a B signal is easy but
separation of a G signal and an R signal becomes imperfect. That
is, a great deal of R light is also absorbed in the whole region of
a PN junction depth at which G light is absorbed. To cope with the
tendency of the imperfect color separation of a G signal and an R
signal, the device structure of FIG. 1 constitutes PDs of two sorts
of R light and B light by superimposing the respective PN junction
surfaces in the depth direction, and arranges PD for G light
independently. Accordingly, the color separation property of each
light wavelength band of RGB can be improved, the occupied area can
be reduced, compared with the case of FIG. 3 where each PD of the
RGB light is dispersed in the plane direction, and simplification
of the semiconductor layer structure can be realized, compared with
the case of FIG. 4 where all of PDs of RGB light are arranged in
the depth direction.
[0092] The high-concentration impurity layer 6 acts so as to pull
in an undesirable dark current, which flows in the surface portion
of a semiconductor region, on the ground of the contamination on
the surface of the semiconductor region due to semiconductor
process, to a common potential or a ground potential to which the
anodes (A_C) of the first to the third photodiode are coupled.
Therefore, the high-concentration impurity layer 6 contributes to
suppress the situation where such a dark current flows into the
cathode and reduces the photoelectric conversion accuracy.
[0093] To cope with the color separation imperfection of a G signal
and an R signal when superimposing the PN junction surfaces of
three sorts of PDs corresponding to the light wavelengths of RGB in
the depth direction, a PD to G light which has a poorest color
separation performance is arranged independently, and the PN
junction surfaces to two sorts of PDs for R light and B light are
superimposed in the depth direction. Accordingly, reduction of the
occupying area is attained. As illustrated in FIG. 21, it is
effective to arrange the superimposed second photodiode (R) and
third photodiode (B), and the first photodiode (G) in a matrix in
the shape of a checkered pattern. According to this arrangement, by
assigning a pixel for every photodiode in a planar arrangement of
the photodiodes arranged in a matrix, and by performing an
interpolation color operation using pixels which adjoin mutually in
a vertical and a horizontal direction, it is possible to detect an
image with the same pixel number as the element number of the array
of the photodiodes arranged in a matrix. The interpolation
operation can be performed very simply and with high accuracy, by
calculating the arithmetic average of the chrominance signals
obtained by the photodiodes corresponding to four surrounding
pixels, as illustrated in FIG. 21, for example.
[0094] The outline is explained about the manufacturing method of
the photoelectric conversion element of FIG. 1. An N.sup.+-type
silicon semiconductor wafer is prepared and an N.sup.--type
semiconductor area is formed in the principal plane by epitaxial
growth. A P-type semiconductor area 1 is formed in the N.sup.--type
semiconductor area by ion implantation and annealing. An N-type
semiconductor area 3 is formed in the P-type semiconductor area 1
by ion implantation and annealing. A P-type semiconductor area 4 is
formed in the N-type semiconductor area 3 by ion implantation and
annealing. Next, an N-type semiconductor area 2 is formed in the
P-type semiconductor area 1 by ion implantation and annealing. An
N-type semiconductor area 5 is formed in the P-type semiconductor
area 4 by ion implantation and annealing. A P.sup.+-type
semiconductor layer is formed thinly (for example, about 0.2 .mu.m
in thickness) as a high-concentration impurity layer 6 over the
surface by ion implantation and annealing.
[0095] FIG. 5 is a circuit diagram illustrating the green
photodiode and a detector circuit of a photoelectric conversion
signal by the green photodiode. FIG. 6 is a plane layout
illustrating planar structure of the circuit of FIG. 5. The symbol
M1 denotes a first transfer MOS transistor of which one of the
source and the drain is coupled in series to a cathode terminal K_G
of a green photodiode PDG. The present first transfer MOS
transistor M1 is coupled to a charge accumulation-output unit ACCG.
The charge accumulation-output unit ACCG accumulates, via the first
transfer MOS transistor M1, the charge information produced by a
current which is induced by photoelectric conversion and flows in
the junction surface JNC_G of the green photodiode PDG, and outputs
the accumulated charge information. The charge accumulation-output
unit ACCG includes: a source follower output MOS transistor M2 of
N-channel type; a selection MOS transistor M3 of N-channel type;
and a reset MOS transistor M4 of N-channel type. The gate of the
source follower output MOS transistor M2 is coupled to the other
one of the source and the drain of the transfer MOS transistor M1,
and the drain of the source follower output MOS transistor M2 is
coupled to a power supply voltage VDD. The selection MOS transistor
M3 selects the output of the source follower output MOS transistor
M2. The reset MOS transistor M4 charges selectively a path from the
gate of the source follower output MOS transistor M2 to the cathode
of the corresponding photodiode. The path from the gate of the
source follower output MOS transistor M2 to the transfer MOS
transistor M1 is configured as a floating diffusion (FD) which has
a comparatively large parasitic capacitance. The reset MOS
transistor M4 supplies the power supply voltage VDD to the charging
node when the reset signal RST is a high level. The selection MOS
transistor M3 is controlled to switch by a selection signal
SEL.
[0096] The imaging cycle using a photoelectric conversion element
is classified roughly into a reset cycle, an exposure cycle, and a
transfer cycle. In the reset cycle, the selection MOS transistor M3
is turned off, and the transfer MOS transistor M1 and the reset MOS
transistor M4 are turned on. As a result, the path from the gate of
the source follower output MOS transistor M2 to the cathode K_G is
charged with the power supply voltage VDD, and an initial charge is
accumulated in the cathode. In the exposure cycle, the MOS
transistor M1, M3, and M4 are turned off, and the photoelectric
conversion is performed to the photodiode PDG. In the next transfer
cycle, the transfer MOS transistor M1 is turned on, and the
photoelectron accumulated in the cathode of the photodiode PDG is
transferred to FD. Then, after the transfer MOS transistor M1
turned off, the selection MOS transistor M3 is turned on. The
detection signal OUT amplified by the source follower output
transistor M2 of which the mutual conductance is controlled by the
voltage of FD is outputted from the selection MOS transistor M3.
Since the transfer MOS transistor M1 is arranged between the charge
accumulation-output unit ACCG and the cathode K_G of the photodiode
PDG, it is possible to suppress that the detection signal OUT taken
out from the charge accumulation-output unit ACCG becomes unstable
under the influence of noises due to exposure.
[0097] FIG. 7 illustrates planar structure of the blue photodiode
PDB and the red photodiode PDR and detector circuits for
photoelectric conversion signals by the blue photodiode and the red
photodiode. The detector circuit for the photoelectric conversion
signal, coupled to the blue photodiode PDB, has a transfer MOS
transistor M11 and a charge accumulation-output unit ACCB. The
charge accumulation-output unit ACCB includes a source follower
output MOS transistor M12, a selection MOS transistor M13, and a
reset MOS transistor M14, and acts similarly as explained in FIG.
5. The detector circuit for the photoelectric conversion signal,
coupled to the red photodiode PDR, has a transfer MOS transistor
M21 and a charge accumulation-output unit ACCR. The charge
accumulation-output unit ACCR includes a source follower output MOS
transistor M22, a selection MOS transistor M23, and a reset MOS
transistor M24, and acts similarly as explained in FIG. 5.
[0098] FIG. 8 illustrates the entire layout of one photoelectric
conversion element constituted by PDG of FIG. 6 and PDB and PDR of
FIG. 7. FIG. 9 illustrates the entire layout of one photoelectric
conversion element in which PDG, PDB, and PDR corresponding to FIG.
3 are dispersively arranged in plane. The occupied area of FIG. 8
is smaller than that of FIG. 9. In FIG. 8, since the charge
accumulation-output units ACCR, ACCG, and ACCB are provided for
every photodiode PDR, PDG, and PDB, it is possible to obtain the
parallel output of the detection signals to the wavelengths of R,
G, and B, by the parallel operation of the charge
accumulation-output units ACCR, ACCG, and ACCB.
[0099] FIG. 10 illustrates a circuit diagram when employing a
charge accumulation-output unit ACCRB provided in common to
photodiodes PDR and PDB. FIG. 11 illustrates the planar structure
of the circuit of FIG. 10. Here, the charge accumulation-output
unit ACCRB includes a source follower output MOS transistor M32, a
selection MOS transistor M33, and a reset MOS transistor M34. The
gate of the source follower output MOS transistor M32 and the
source of the reset MOS transistor M34 are coupled in common to a
floating diffusion FD of the photodiode PDR and a floating
diffusion FD of the photodiode PDB. In photoelectric conversion
operation, the transfer cycle using the charge accumulation-output
unit ACCRB is performed by time sharing of the transfer cycle of
PDB and the transfer cycle of PDR. The occupied area of FIG. 11 is
smaller than that of FIG. 7.
[0100] FIG. 12 illustrates a circuit diagram when employing a
charge accumulation-output unit ACCRBG provided in common to
photodiodes PDR, PDB, and PDG. FIG. 13 illustrates the planar
structure of the circuit of FIG. 12. Here, the charge
accumulation-output unit ACCRBG includes a source follower output
MOS transistor M42, a selection MOS transistor M43, and a reset MOS
transistor M44. The gate of the source follower output MOS
transistor M42 and the source of reset MOS transistor M44 are
coupled in common to the respective floating diffusions FDs of
photodiodes PDR, PDG, and PDB. In photoelectric conversion
operation, the transfer cycle using the charge accumulation-output
unit ACCRBG is performed by time sharing of the transfer cycle of
PDB, the transfer cycle of PDR, and the transfer cycle of PDG. The
occupied area of FIG. 13 is smaller than that of FIG. 11.
[0101] FIG. 14 illustrates a still-more-detailed vertical section
structure of a photoelectric conversion device to which a transfer
MOS transistor is added. The symbol 10 denotes a gate oxide layer
and the gate (GT) of the transfer MOS transistor is formed through
this gate oxide layer 10. Over that, a light shielding film 12 and
a floating diffusion FD are formed through an interlayer insulation
film 11.
[0102] All or a part of the transfer MOS transistors M1, M11, and
M21, the selection transistors M3, M13, M23, M33, and M43, the
source follower output transistors M2, M12, M22, M32, and M42, and
the reset MOS transistors M4, M14, M24, M34, and M44, may employ a
bulk MOS transistor of FIG. 20 and noise can be reduced. Namely,
the bulk MOS transistor has structure in which the diffusion layer
whose impurity concentration is deeper than in the channel forming
layer is provided in the interface under the gate, and a carrier is
made to pass only the bulk, without passing the interface.
Therefore, noise can be reduced.
[0103] FIG. 15 illustrates another outline cross-section structure
of a photoelectric conversion device which configures a solid-state
imaging device. It is assumed that the semiconductor silicone
substrate is of N.sup.+-type, for example, and an N.sup.--type area
is epitaxially grown on the semiconductor silicone substrate. In
the N.sup.--type area, many photoelectric conversion elements are
formed at a predetermined pitch in a matrix.
[0104] The photoelectric conversion element includes a first
semiconductor region 1 of a first conductive type, for example, a P
type, formed in the N.sup.--type area, and a second semiconductor
region 2 and a third semiconductor region 3A of a second conductive
type, for example, an N type, which are arranged in the first
semiconductor region 1. Although the second semiconductor region 2
is shaped like a pillar same as in the above, the third
semiconductor region 3A is shaped like a stepped pillar with a step
attached at the intermediate point in the depth direction, the
cross-sectional area of the deep portion is formed larger than the
cross-sectional area of the shallow portion, and the upper and
lower surfaces of the deep part of the third semiconductor region
3A contact the first semiconductor region 1. In short, the first
semiconductor region and the fourth semiconductor region in FIG. 1
are formed in a unified shape. A semiconductor region 5A of the
same N type as the fifth semiconductor region is formed in the
position which overlaps with the deep part of the third
semiconductor region 3A in the first semiconductor region 1. A blue
photodiode is configured by the junction surface JNC_B between the
semiconductor region 5A and the first semiconductor region 1. A red
photodiode is configured by the junction surface JNC_R between the
third semiconductor region 3 and the first semiconductor region 1.
A green photodiode is configured by the junction surface JNC_G
between the first semiconductor region 1 and the second
semiconductor region 2. Other constitution is the same as the
photoelectric conversion element explained in FIG. 1, therefore,
the detailed explanation thereof is omitted. The photoelectric
conversion element of this structure also functions similarly as
the photoelectric conversion element described above.
[0105] FIG. 16 illustrates yet another outline cross-section
structure of a photoelectric conversion device which constitutes a
solid-state imaging device. The photoelectric conversion element
illustrated in FIG. 16 is different from the structure of FIG. 14
in the point that the cathode terminal K_B of the blue photodiode
is arranged next to the cathode terminal K_G of the green
photodiode. Since other constitution is the same as that of FIG.
15, the detailed explanation thereof is omitted. The photoelectric
conversion element of this structure also functions similarly as
the photoelectric conversion element described above.
[0106] FIG. 17 illustrates circular vertical section structure in
both ends of a third to a fifth semiconductor region, as another
example of the shape of the laminated semiconductor regions. An
antireflection film 20 is formed over the surface of the
semiconductor region, and a light shielding film 22 with an opening
21 formed above the fifth semiconductor region 5 is provided. The
third semiconductor region 3 to the fifth semiconductor region 5
are formed so as to have the second depth (DP_R) and the third
depth (DP_B) in the direction of a light diffracted outwards at a
periphery portion of the opening 21, similarly as in the direction
of a light entering substantially perpendicularly to the opening
21. Even when the incident light is diffracted at the periphery of
the opening 21 and the direction of the light becomes slanting, the
light path length cannot change greatly compared with the
perpendicular light; therefore, the accuracy deterioration of the
wavelength separation or the color separation in one photodiode can
be suppressed. Although not illustrated in particular, also as for
the PN junction surface JNC_G of the green photodiode, both ends of
the second semiconductor region 2 may be preferably shaped
circularly at the second depth. The antireflection film 20
functions also as a protective film (a passivation film) and may be
formed by depositing at least one of SiO.sub.2, SiON, and SiN, by a
molecule deposition method. The light shielding film 22 may be
formed by applying resin of a pigment including a black dye, for
example, by a photoengraving construction method.
[0107] FIG. 18 illustrates vertical section structure which
combines a lens with a red photodiode and a blue photodiode
superimposed in the vertical direction, as another example in which
the lens is arranged in front of a photoelectric conversion
element. Over the upper surface of the antireflection film 20, a
light shielding film 22 with an opening 21 formed at the upper
position of the fifth semiconductor region 5 is formed. A concave
lens 23 formed by translucent material is disposed in the opening
21, and a convex lens 24 formed by translucent material is disposed
over the concave lens 23. The convex lens 24 converges the incident
light to the photodiode and improves the condensing of light. The
concave lens 23 changes the condensed light to a parallel light,
and make the light enter substantially perpendicularly to the
photodiode. With the combination of the concave lens 23 and the
convex lens 24, excellent color separation performance as well as
condensing of light can be obtained.
[0108] The concave lens 23 can be formed in the following way. That
is, a light-transparent resin is applied over the light shielding
film 22 and the antireflection film 20, the present resin in the
opening 21 (the light-sensitive section) is gouged out by a
photoengraving construction method; subsequently, roundness is
given to the shape to form a concave by melting by heat, and the
concave lens 23 is formed. By applying a light transparent resin
over the upper part of the present concave lens 23, a planarizing
layer (a planarizing film) 25 is made. The convex lens 24 can be
formed in the following way. That is, a light transparent resin is
applied over the upper part of the planarizing layer 25, the
applied resin is arranged to a shape of a cylinder or the like at a
light-sensitive section by a photoengraving construction method;
subsequently, roundness is given to the shape to form a convex by
melting by heat, and the convex lens 24 is formed.
[0109] The light shielding films 22 may be formed from metal
silicide films, such as tungsten silicide (WSi), molybdenum
silicide (MoSi), and titanium silicide (TiSi), or it may be formed
from metal films, such as tungsten (W), molybdenum (Mo), titanium
(Ti), and wiring metal aluminum (Al). Sputtering or CVD is used to
form the light shielding films 22. When forming the light shielding
film from a metal silicide film or a metal film, boron-phosphorus
silicide glass (BPSG) may be formed over the surface of the light
shielding film by CVD.
[0110] FIG. 19 illustrates the system configuration of an imaging
device. The symbol 30 denotes a solid-state imaging device which is
formed, for example, by a CMOS integrated circuit manufacturing
method. The solid-state imaging device 30 is constituted by
photoelectric conversion elements arranged in the shape of an array
on one semiconductor substrate. An optical image (light) enters
into the solid-state imaging device 30 through a lens 31, a
aperture diaphragm 32, an infrared ray cut filter and optical LPF
33. The solid-state imaging device 30 performs the color conversion
of RGB to the incident light, and outputs a detection signal (image
signal). An analog front-end unit to digitize the image signal
includes CDS (correlated double sampling circuit) 34, GCA (gain
control amplifier) 35, and ADC (analog-digital converter) 36. DSP
(digital signal processing processor) 38 performs digital signal
processing to the output of ADC 36 to generate image data. The
image data is displayed by DISP (LCD display) 39, and also stored
in a flash memory (FLASH) 41 or the like through a media interface
(MDAI/F) 40. It is also possible to output the image data to an
external PC (personal computer) etc. through an external interface
(EXI/F) 42. A data processor (MCU) 43 controls the entire system. A
timing generator (TGEN) 45 performs timing control to the analog
front-end unit. An aperture diaphragm driver (APDRV) 46 drives the
aperture diaphragm 32. A lens driver (LZDRV) 47 performs focus
control of the lens 31.
[0111] In the above, the invention accomplished by the present
inventors has been specifically explained based on the embodiments.
However, it cannot be overemphasized that the present invention is
not restricted to the embodiments, and it can be changed variously
in the range which does not deviate from the gist.
[0112] For example, a P-type substrate may be used as the silicon
semiconductor substrate. The MOS transistor which constitutes the
charge accumulation-output unit is not restricted to an N-channel
type, but may use a P-channel MOS transistor in part. It is
needless to say that the PN junction depth of each photodiode of
RGB is not restricted to the explanation given above but it can be
changed suitably.
INDUSTRIAL APPLICABILITY
[0113] The present invention is widely applicable to a solid-state
imaging device and a photoelectric conversion element which
constitute an image input device or an imaging device, such as a
video camera, a digital still camera, and a scanner.
* * * * *